[00:28:44] --- lars.malinowsky has left [00:29:05] --- Russ has left: Disconnected [00:56:14] --- lama has become available [01:57:33] --- abo has left [02:03:04] --- abo has become available [02:08:49] --- haba has become available [03:13:50] --- haba has left [03:14:23] --- haba has become available [03:18:14] --- abo has left [03:29:37] --- abo has become available [04:13:03] --- Simon Wilkinson has become available [05:09:51] --- Simon Wilkinson has left [07:34:55] --- reuteras has left [07:39:57] --- deason_gmail has become available [07:48:54] --- haba has left [07:49:32] --- haba has become available [08:20:26] --- lama has left [08:36:17] --- ksumner has become available [09:24:44] --- haba has left [10:16:20] --- abo has left [12:47:40] what's the best way to reach mdionne? [12:48:45] depends if you wanna spring for a flight. if not, probably buses. [12:50:05] 1.6.0pre5 with the 4 line patch he sent me privately still panics when bypass is enabled, and in the same manner [12:50:46] i take it you replied privately and ae now being impatient? ;) [12:51:14] we spoke via jabber, and it seems that he's not online right now [12:51:58] probably email or wait. [12:52:10] you have a backtrace? [12:52:58] I have a screenshot of the panic from the console [12:53:09] drop in afs, or if you can't, shadow@gmail [12:53:50] /afs/bx.psu.edu/user/phalenor/public/1.6.0pre5_cbp_panic.jpg [12:54:02] replace the 5 with a 4, and you get almost the same thing [12:57:33] --- ksumner has left [13:03:09] --- lars.malinowsky has become available [15:49:01] > 1.6.0pre5 with the 4 line patch he sent me privately still panics >when bypass is enabled, and in the same manner got any way to get details from the other threads? [15:49:53] that's it unfortunately [15:50:18] machine locks up hard, can only reset it, unless you know of some trick with ESX [15:51:48] not a clue. wanted to know if it was readpage or readpages [16:03:12] --- mdionne has become available [16:07:37] phalenor: are there any afs related messages preceding the oops in the syslog [16:09:32] the patch I gave you would not prevent the panic - it prints out certain variables if they are outside the expected bounds [16:17:56] --- lars.malinowsky has left [16:31:19] --- deason_gmail has left [17:43:52] --- jaltman/FrogsLeap has left: Replaced by new connection [17:43:54] --- jaltman/FrogsLeap has become available [17:48:28] --- jaltman/FrogsLeap has left: Disconnected [17:48:44] --- jaltman/FrogsLeap has become available [18:07:07] mdionne: aha! [18:07:16] afs: iovoff out of range (bypass): 0, iov_len is 0 [18:11:41] > iov_len is 0 yeah, i kinda wondered if that was true [18:12:28] interesting, so we have an iov of length 0. [18:13:29] i don't know what you gave him, where's that printf? [18:13:51] assuming that's normal, maybe we can just skip the copy and move on [18:14:05] + printf("afs: iovoff out of range (bypass): %d, iov_len is %d\n", iovoff, rxiov[iovno].iov_len); [18:14:37] yeah, where tho [18:14:46] like, which function did you add that in [18:14:51] just before the first memcpy [18:14:56] well, the variable name tells me [18:14:57] --- jaltman/FrogsLeap has left: Disconnected [18:15:16] in afs_NoCacheFetchProc [18:16:51] yeah. looking now. [18:20:57] kind of curious what iovno was. [18:21:30] --- jaltman/FrogsLeap has become available [18:22:19] iovno is the current iov we're using [18:22:34] sure. i mean, what's the value? or did i miss that? [18:22:40] i deleted too much stuff [18:23:08] it wasn't printed in the printf [19:10:02] --- rra has left: Disconnected [19:12:14] --- rra has become available [19:13:14] Okay, that was a new one. I did an fs setacl on a directory, followed shortly by an fs listacl. The fs listacl hung but the shell was still responsive (pressing Enter moved the screen) for about a second or two, and then the system locked up with the same symptoms as the previous 1.6.0pre4 problem. [19:13:32] This is still with 1.6.0pre4 on 2.6.38 Debian with Simon's patch. [19:30:16] --- mdionne has left [19:41:39] How do we feel about posix_fadvise()? [19:41:54] it hasn't done anything to us lately? [20:21:05] --- rra has left: Disconnected [20:48:05] --- Russ has become available [22:28:14] --- lars.malinowsky has become available [23:19:45] --- reuteras has become available [23:21:24] --- lars.malinowsky has left [23:54:04] --- abo has become available